Product Experimentation for Growth
summary
Gives a playbook: start with lots of ideas, run rapid experiments using prototypes and split traffic A/B tests, and limit work to a few track 1–3 core OKRs that must show no harm. Cites WD‑40’s 39 failed attempts before success and Amazon’s marketplace and Prime as experiments that cannibalized but validated demand. Emphasizes capture-and-share learning, compare test vs control cohorts, and prioritize only experiments that move customer usage, retention, revenue or NPS.
AI-written summary · about
chapters (5)
- 0:00Introduction to growth
- 1:54Why experiment matters
- 3:37Culture and leadership mindset
- 8:11Approach to experimentation
- 11:40Learning from experiments
related talks
01h 03mExperimentation in Data Science
34mHypothesis to Impact: Mastering the Art and Science of Controlled Experiments
22mPowering the Next Generation of Products
18mFast Scale
34mMaximizing marketplace experimentation: switchback design for small samples
40mDesigning Data Driven Projects: Beyond ML Experimentation
31mA data innovation playbook
26mML Product Experiments at Scale